Twi
Answers are digits and aane / daabi, recognised by Vomo’s own on-device model. Nothing leaves the phone to understand them, so it works on a weak connection.
Ghana · MTN MoMo · Twi
“Fa fifty cedis kɔma Ama.”
That sentence is the whole interaction. Vomo reads back who and how much, waits for one confirmation, then opens MTN’s *170# menu and types every step on your behalf — the recipient, the amount, the reference and the PIN.
Seven entries, one spoken sentence, no taps.
Everything Vomo needs is decided before it dials. Once the menu opens there is no time to ask you anything.
Say “Hey Vomo”, or tap Kasa. Vomo can also listen while it’s closed, so the phone never has to be unlocked and navigated first.
Speak naturally in Twi or English. Phone numbers go in one digit at a time, in either language, and you can take the last one back.
Vomo says the name and the amount out loud and shows them large on screen. If it wasn’t sure what it heard, it asks instead of guessing.
Say 1 to send or 2 to stop. This is the last checkpoint before money moves, so a spoken “1” needs higher confidence here than anywhere else.
About seven entries in roughly twenty seconds, with the progress read aloud. You end up on MTN’s own receipt — recipient, new balance, transaction ID.
Set in Settings, under “How to talk to Vomo”. Both wake on the same phrase.
Answers are digits and aane / daabi, recognised by Vomo’s own on-device model. Nothing leaves the phone to understand them, so it works on a weak connection.
The same questions take ordinary answers — “check my balance”, “yes”. A phrase Vomo can’t use is asked again twice, then it goes back to waiting.
Phone numbers are the exception: in both modes Vomo’s own model hears the digits, with the microphone held open until all ten are in.
MTN’s menu is English-only, text-heavy and on a timer. That is a poor fit for a lot of people who depend on it every day.
Audio is a first-class output, not an add-on. Every screen also works heard and through TalkBack, for people with low vision or limited literacy.
Twi is the default, not a translation layer bolted on afterwards. Amounts are the one thing people say in English, so Vomo expects that inside a Twi sentence.
The session lasts twenty to thirty seconds. Vomo decides everything up front and then types faster than anyone can, so the timer stops being your problem.
Real money moves the moment MTN accepts the PIN, and there is no undo. The rules around it are deliberate.
Not simulated and not mocked. Every line below has been run end to end on a live MTN SIM, on a real phone.